Mozilla slips an extra beta into Firefox schedule

By Gregg Keizer, Computerworld (US)

Mozilla is to add a third beta to the development schedule for Firefox 3.1. The move will enable the company to get a better handle on remaining bugs and give several new features, including a faster JavaScript engine and a private browsing mode, more testing time, the company's browser director said yesterday.

Previous schedules published by Mozilla had limited Firefox 3.1 to only two betas before moving to a release candidate.

In a long post to the mozilla.dev.planning forum, Mike Beltzner, the director of Firefox, said that Beta 3 was necessary to get a feel for the severity of the remaining bugs and an idea of how long it will take developers to eradicate them. In addition, another beta will give more exposure to features landing in the browser only as of Beta 2, which has not yet been released.

Beltzner named several of Firefox 3.1's high-profile additions, including the new "TraceMonkey" JavaScript rendering engine and the so-called "porn mode" feature, dubbed "Private Browsing Mode" by Mozilla, among those that would benefit from more testing.

Beta 3 is not a done deal, Beltzner noted in a follow-up email, but he is confident that developers would approve the plan.

"We're never comfortable declaring new milestones by fiat, but I expect that there won't be any opposition to the plans for a third beta at today's meeting," he said. "I'd say that it's very likely at this point."

So far, Mozilla has shipped only Beta 1, which was released six weeks ago, although Beta 2 should be available in early December, perhaps as soon as the end of next week. A schedule for the third beta has not been set, but Beltzner said Mozilla would likely declare a "code freeze" - a milestone after which changes are either forbidden outright or tightly restricted - in early January.

Beltzner stressed that the extra beta wouldn't delay the final version of Firefox 3.1.

"We believe we can do this without major impact to our shipping schedule," he said.

"It's more a matter of inserting another public consultation milestone than it is about slipping, per se."

Mozilla is traditionally leery of committing to final ship dates - like other developers it typically says it launches products when they're ready, not on a timetable - but previously it had said it was shooting for a late 2008 or early 2009 window.

Yesterday, however, Beltzner said that Firefox 3.1 is "still looking at late in Q1 2009 for final delivery".

In a status meeting last week, Mozilla also decided to retract a revamped Ctrl-Tab tab-switching feature it had originally slated for Firefox 3.1.

The enhancement, which was based on an already-available Firefox add-on, showed users thumbnails when they cycled through open tabs, and switched between current and last-viewed tabs rather than simply moving to the next tab to the right.

Like many of the features that made it into Firefox 3.1, it was initially set for Firefox 3.0, but had slipped out of that earlier update.

"This is something that our development community is getting used to," Beltzner said Tuesday in another email.

"As part of our new effort to try and increase the pace of our releases, we're coming to terms with the fact that new user-facing features sometimes need a lot of trial-and-error to get feeling right."

The newest attempt at redesigning tab switching, said Beltzner, didn't "feel quite right" to either Mozilla's user interface team or outside testers; the feature was pulled as a result.

"If we can get it right, we can see if we still have time for its inclusion," he added.

"If not, it can wait until the next release."

Mozilla was the second browser developer in the last week to claim it is still on schedule. Last Thursday, Microsoft laid out the next step for Internet Explorer 8 (IE8); Monday, a Microsoft director of product management denied that IE8's schedule had slipped in an interview with the Seattle Post-Intelligencer.

"It's not a slipped schedule," Matthew Lapsen, a director of Windows product management, told the newspaper. "We release based on product quality, not dates."

Some Microsoft officials, however, had alluded to a late-2008 ship date for IE8 several months ago.

Original here

Micron promises supercharged 1 GB/s SSD

By Wolfgang Gruener

Chicago (IL) – Chip manufacturer has demonstrated what is, at least to our knowledge, the fastest solid state disk drive (SSD) demonstrated so far. A demo unit shown in a blurry YouTube video was hitting data transfer rates of 800 MB/s and can expand to apparently about 1 GB/s. The IO performance is about twice of the best performance we have seen to date.

Ok, we have not actually seen the exact numbers, since the YouTube video is too blurry, but if we believe the claims made my Micron staffers, then we might see a dramatic increase in SSD performance in the not too distant future.

Posted on Micron’s new blog, the video showcases a 2-processor, eight-core Intel Xeon PC with two SSDs The difference to your average SSD is that these drives are not connected via a PATA or SATA interface and therefore are not limited to the bandwidth limit of 300 MB/s of SATA II. Instead, the drives have PCIe interfaces and, according to Micron, include flash data management enhancements.

During the short demonstration, Micron’s Joe Jeddeloh claims that the two cards hit a data throughput of about 800 MB/s and about 150,000 to 160,000 random IOPS. He also showed a flash PCIe card that combined the two cards in one device with 16 flash channels. This card will be hitting a bandwidth of 1 GB/s and “at least 200,000 IOPS,” he said. For comparison reasons, the fastest enterprise SSDs available from big manufacturers today hit about 250 MB/s and about 30,000 IOPS.

The fastest hard drive you can buy today, WD’s Velociraptor, clocks in at about 100 MB/s.

The PCIe concept of SSDs has been shown before, most notably by Fusion IO, which hit about 100,000 IOPS about a year ago. The company recently announced a “consumer version” of its PCIe card, which will offer data throughput of 500 – 700 MB/s and about 50,000 IOPS for about $1000.

We have no idea what the Micron drive will cost, but it does not look cheap. Jeddeloh noted that the 1 GB/s drive will be available “soon”.

Original here

The Electronic Frontier Foundation Doesn't Like Apple's Attitude

by Sam Dean

Who knows why many open source users are also Macintosh users, but I've noticed a correlation there for years. Maybe it's because open source, like the Mac, rings of rebellion against the status quo. Nevertheless, if you think the love always flows in two directions, check out this post from the Electronic Frontier Foundation. According to the EFF, Apple's lawyers recently put the kibosh on an online discussion of open source software, dubbed iPodhash, which the lawyers perceived as designed to circumvent aspects of Apple's FairPlay digital rights management (DRM), used in iTunes. (The original discussion is removed.) The key word there is "discussion"--there was no fully realized application that actually cracked Apple's encryption.

As the Electronic Frontier Foundation points out:

"At the heart of this is the iTunesDB file, the index that the iPod operating system uses to keep track of what playable media is on the device. Unless an application can write new data to this file, it won't be able to 'sync' music or other content to an iPod. The iTunesDB file has never been encrypted and is relatively well understood. In iPods released after September 2007, however, Apple introduced a checksum hash to make it difficult for applications other than iTunes to write new data to the iTunesDB file, thereby hindering an iPod owner's ability to use alternative software (like gtkpod, Winamp, or Songbird) to manage the files on her iPod."

Apple's original checksum hash, an encryption scheme, was cracked within three days of its release. However, Apple has recently updated the checksum hash for the iPod and iPod Touch, and that piece of code obfuscation has not been hacked yet.

Long story short, a discussion of how to crack Apple's new checksum hash using open source software called iPodhash took place on a wiki dubbed Bluwiki. While nobody posted an actual application for cracking it, there was some illustrative code. You can view the communications that Bluewiki has posted regarding taking the discussions down here. There, the Digital Mellenium Copyright Act (DMCA) is cited as protecting Apple's FairPlay DRM scheme, because it is considered "anti-circumvention technology."

Attorneys at the Electronic Frontier Foundation are crying foul. "If Apple is suggesting that the DMCA reaches people merely talking about technical protection measures, then they've got a serious First Amendment problem," they write.

I have to agree. Talk is talk. Encryption and other forms of code obfuscation are communally practiced throughout the worlds of proprietary and open source software. Also, encryption predates personal computers and Apple by many years. Where would security software end up if people were barred, directed by lawyers, from ever discussing methods of disguising code?

Original here